Gov Corruption and Failure

SamSpade

SamSpade

Well-Known Member
PREMO Member
GURPS said:
🔥
🔥
🔥


In the latest electrifying development that will also probably not shock you, DOGE —for the last week slogging through the SSA’s septic systems of archaic paper and antique Cobol software — is beginning to report that the OIG canaries in the SSA mineshaft were, if anything, woefully underreporting the problem.

Yesterday, apparent DOGE mouthpiece Elon Musk posted a simple database count of active Social Security Numbers by age range. The dumbfounding chart was a political hydrogen bomb, and it speaks for itself:

image 3.png
Click to expand...
Having read any number of reddits and substacks, I'm convinced this is not due to some "default setting" for a date field.

For one thing, IF that were true - there'd be a HUGE cluster around the 150 exactly date - and there's not. For another - COBOL doesn't have a "date" data type - just character. Which means, if you leave it blank - it's blank. If you put in all zeroes, it has all zeroes. Some other languages - Java, Python, SAS - DO have date data types, and nulls translate to January 1, 1970. Moreover - COBOL was invented here - why would we set a default setting to some event in FRANCE where they standardized the meter? And not January 1 but some date in the middle of the year?

No - this is a case of a meme that got wings from people repeating it. It's NOT because COBOL is old and has a dumb default. I KNEW this wasn't the case, but I haven't coded in COBOL since Y2K. People are repeating this, and they don't know COBOL - and idiotic to call others dumbasses about a language they don't know, either. People subtracted 150 years from today's year (BTW - Elon has also said other years) and manufactured some BS about 1875.

HOWEVER - it IS possible that local systems have made policy about how data is recorded. This wouldn't be a failing of COBOL - but of the Social Secuirty Administration. Once Y2K hit, lots of systems still using COBOL had to adapt for recording dates. But at that point, 1875 would have been 125 years in the past - there's simply no logical reason whatsoever to use THAT as a base date.

Even if you try to include early beneficiaries of SS, it still doesn't work. You have to have 40 quarters of work and the program came out in 1935 (although back then, it was in years, and the first checks came out in 1940). The first recipient was actually born in 1874. However, Ida Mae Fuller died fifty years ago.

My opinion? You cannot have an OLD birthdate going back 180-200+ years, no matter how you analyze it. This is data corruption. This is not a failure to retire old numbers. This is EXTREMELY BAD maintenance of data. People can mock how Elon put the drama out there - but there's no way around it - data is poorly managed. VERY poorly.

Here's a good page where the matter is discussed --
retrocomputing.stackexchange.com

Did missing/corrupt dates in COBOL default to 1875-05-20?

There is a claim circulating, e.g. here on Mastodon: [...] in COBOL, if a date is missing [or "corrupt"] the program defaults to 1875 [...] The claim is sometimes "enhanced" by
retrocomputing.stackexchange.com retrocomputing.stackexchange.com
 
GURPS

GURPS

INGSOC
PREMO Member

Defense Secretary Pete Hegseth says Biden’s IRS targeted him for an audit​



“Of course the outgoing Biden IRS rushed an ‘audit’ of the incoming SecDef. Total sham,” Hegseth wrote on his official X account Monday, including a photo of an IRS document addressed to him and his wife.

The document said the couple owes the government $33,558.16, which needs to be paid immediately to avoid further penalties.

It’s unclear what year or years the document is addressing.

“The party of ‘norms’ and ‘decency’ strikes again. We will never back down,” Hegseth added in his tweet.
 
stgislander

stgislander

Well-Known Member
PREMO Member
SamSpade said:
Having read any number of reddits and substacks, I'm convinced this is not due to some "default setting" for a date field.

For one thing, IF that were true - there'd be a HUGE cluster around the 150 exactly date - and there's not. For another - COBOL doesn't have a "date" data type - just character. Which means, if you leave it blank - it's blank. If you put in all zeroes, it has all zeroes. Some other languages - Java, Python, SAS - DO have date data types, and nulls translate to January 1, 1970. Moreover - COBOL was invented here - why would we set a default setting to some event in FRANCE where they standardized the meter? And not January 1 but some date in the middle of the year?

No - this is a case of a meme that got wings from people repeating it. It's NOT because COBOL is old and has a dumb default. I KNEW this wasn't the case, but I haven't coded in COBOL since Y2K. People are repeating this, and they don't know COBOL - and idiotic to call others dumbasses about a language they don't know, either. People subtracted 150 years from today's year (BTW - Elon has also said other years) and manufactured some BS about 1875.

HOWEVER - it IS possible that local systems have made policy about how data is recorded. This wouldn't be a failing of COBOL - but of the Social Secuirty Administration. Once Y2K hit, lots of systems still using COBOL had to adapt for recording dates. But at that point, 1875 would have been 125 years in the past - there's simply no logical reason whatsoever to use THAT as a base date.

Even if you try to include early beneficiaries of SS, it still doesn't work. You have to have 40 quarters of work and the program came out in 1935 (although back then, it was in years, and the first checks came out in 1940). The first recipient was actually born in 1874. However, Ida Mae Fuller died fifty years ago.

My opinion? You cannot have an OLD birthdate going back 180-200+ years, no matter how you analyze it. This is data corruption. This is not a failure to retire old numbers. This is EXTREMELY BAD maintenance of data. People can mock how Elon put the drama out there - but there's no way around it - data is poorly managed. VERY poorly.

Here's a good page where the matter is discussed --
retrocomputing.stackexchange.com

Did missing/corrupt dates in COBOL default to 1875-05-20?

There is a claim circulating, e.g. here on Mastodon: [...] in COBOL, if a date is missing [or "corrupt"] the program defaults to 1875 [...] The claim is sometimes "enhanced" by
retrocomputing.stackexchange.com retrocomputing.stackexchange.com
Click to expand...
Then the big question is are any checks being mailed out because of that bad data.
 
SamSpade

SamSpade

Well-Known Member
PREMO Member
stgislander said:
Then the big question is are any checks being mailed out because of that bad data.
Click to expand...
From what I've read - unknown but unlikely. Musk said the records show old dates, and death date set to false - leading him to joke that there are VAMPIRES getting social security.

Also - he wasn't the first to find this - it was first picked up during an internal audit a few years ago.
 
GURPS

GURPS

INGSOC
PREMO Member
SamSpade said:
Having read any number of reddits and substacks, I'm convinced this is not due to some "default setting" for a date field.
Click to expand...

Today's Blog Following up .... follow the link for the rest


☕️ FREEDOM FRIED ☙ Tuesday, February 18, 2025 ☙ C&C NEWS 🦠​



I’ll begin with an apology. We must discuss Social Security a little more. The Deep State’s firefighting bots were out in force yesterday, desparately trying to extinguish the flames of righteous outrage kindled by DOGE’s Zombie Army disclosures.

image 8.png

The distraction that nearly everyone focused on was the threshold question of whether the Zombie Army (people older than 100+ lacking dates of death in the Social Security System) were still receiving direct Social Security payments. The simple answer is we don’t know for sure but it doesn’t matter. The almost-certainly deliberately over-complicated answer is the kind of thing that gives you a rage-induced migraine.

But we shall strain for clarity. Let’s begin with yesterday’s most cited sentence from the OIG’s 2023 Report:

image 5.png

Hecklers focused on the first highlighted sentence above, wherein the OIG apparently admitted the Agency assured him no direct SSA payments were being made to most of the 18.9. million in the 100+ cohort. Accepting that as true for argument’s sake, crater-sized holes remain.

For example, SSA said “almost none” currently receive SSA payments, which is not the same as none. The number of zombies having checks scattered over their gravesites remains unquantified. A better question is: do they even know? Or were they just guessing? If they did know how many there were —“almost none”— then they must also know who they were, and so they could stop the payments. But apparently not.

Or, for another example, how confident are they that deceased people under 100 aren’t still receiving payments? It was never studied.
 
GURPS

GURPS

INGSOC
PREMO Member

Documents Show How Biden’s State Dept Ignored Warnings About Funding Hamas-Linked UN Organization​


In an April 2021 memo to the secretary, a Blinken lieutenant detailed incidents and arguments against restoring the funding.

Some were relatively prosaic: UNWRA had made “little to no progress” on U.N. financial reporting requirements, and a 2020 financial mismanagement investigation resulted in dismissal of three senior officials—nothing that would make a Beltway bureaucrat withhold American taxpayer money.

But there were also the “education” materials UNWRA provided for Palestinian children, which refer to “jihad” and Israeli “occupation.” Such terms were “in line with U.N. principles … but are viewed as inappropriate by some other audiences.” Other audiences can be so unreasonable.

And there were the “violations of UNRWA’s neutrality policies” 247 of them in 2019-20 alone. These ran the gamut from “slogans, graffiti, or other imagery on the inside or outside of UNRWA facilities” to “armed incursions, armed incidents, unauthorized use of UNRWA installations, weapons, and tunnels.” The memo mentions “three different occasions since the last U.S. contribution [2018], a cavity or tunnel was found under an UNRWA installation.”

U.N. inspectors found things like hand grenades and “military vests” in UNWRA facilities, including schools, a health center, and an aid distribution center.

You have to wonder what, if anything, would constitute a deal-breaker for Blinken’s team? Any sober, morally serious analysis wouldn’t even consider dumping American money into the UNRWA cesspool. Instead, Blinken’s staff batted around caveats and excuses, noting how nobly UNRWA had struggled on without U.S. largess.

And the agency had “long-standing staff regulations outlining the neutrality, integrity, and impartiality required.” It had processes in place and “conducts investigations into cases of alleged wrongdoing by staff or beneficiaries, including involvement in militant activities, as well as possible breaches of U.N. neutrality principles.” And whenever anybody found a terrorist tunnel in a UNRWA property, “UNRWA has been very transparent with the Israeli Ministry of Defense and with the United States.” Stand-up guys.

It is difficult to see why the Biden administration was in such a hurry to restart funding to UNWRA. The Palestinians in Gaza were poor, no doubt, but there was no humanitarian emergency until Hamas provoked war with Israel.
 
GURPS

GURPS

INGSOC
PREMO Member
SamSpade said:
Having read any number of reddits and substacks, I'm convinced this is not due to some "default setting" for a date field.
Click to expand...



x.com

twitter.com twitter.com


The "Social Security COBOL lies 1875 date problem" refers to a widely discussed claim that the U.S. Social Security Administration’s (SSA) database, which relies on the COBOL programming language, has a flaw causing records to show people born in 1875—making them 150 years old in 2025—as active beneficiaries. This idea gained traction recently, notably after Elon Musk highlighted supposed evidence of "massive fraud" in the SSA, pointing to recipients over 150 years old. The narrative suggests this stems from COBOL’s handling of dates, with 1875 as a default or "epoch" date for missing birth data. Let’s break this down.

COBOL, a programming language from the 1950s, is indeed still used by the SSA and other government systems for its reliability in handling large-scale transactional data. Unlike modern languages, COBOL doesn’t have a built-in date type, so dates are typically stored as numeric or text fields (e.g., "YYYYMMDD"). How missing dates are handled depends entirely on how programmers design the system—COBOL itself doesn’t impose a default like 1875. The notion that 1875 is a universal COBOL "epoch" (a reference point like January 1, 1970, in Unix) is a misconception. Some have linked this to the ISO 8601 standard, which mentions May 20, 1875 (the date of the "Convention du Mètre" in Paris), but that standard, first published in 1988, postdates the SSA’s core systems (established in the 1960s) and isn’t a COBOL requirement.

So, does the SSA use 1875 as a default for missing birth dates? There’s no definitive public evidence from the SSA confirming this. Critics argue that if this were true, you’d see a spike of records clustered at exactly 150 years old (born May 20, 1875), but Musk’s reported data showed millions of records spread across ages 120 and up, not a single-point anomaly. This suggests something broader—possibly data errors, unupdated death records, or misinterpretation of the database—rather than a COBOL-specific "1875 glitch." For context, the SSA’s Master Beneficiary Record, managing payments since 1962, would have dealt with people born in the late 19th century legitimately (e.g., an 87-year-old in 1962 was born in 1875), but they wouldn’t still be alive or receiving benefits in 2025.

The SSA does have known issues with old data. A 2023 Inspector General report found 18.9 million records of people born in 1920 or earlier weren’t marked as deceased, but 98% of those over 100 weren’t receiving benefits. The agency also stops payments at age 115 automatically since 2015, undermining claims of rampant payouts to 150-year-olds. Experts like Alicia Munnell from Boston College have dismissed the scale of fraud Musk implied (e.g., 40% of beneficiaries being over 100) as implausible—overpayments are real but hover around 1% of benefits.

The "1875 problem" seems more like a misunderstanding amplified online than a COBOL lie. It’s possible some SSA applications use 1875 as a placeholder (a choice made by programmers, not COBOL), but without official specs, that’s speculative. More likely, Musk’s team misread legacy data quirks—perhaps NULL fields, unverified birth dates, or survivor benefits tied to old records—as fraud. The real issue isn’t COBOL’s "lies" but aging systems struggling with data integrity, a problem modernization could address better than finger-pointing at a 60-year-old language.
 
SamSpade

SamSpade

Well-Known Member
PREMO Member
GURPS said:
The "1875 problem" seems more like a misunderstanding amplified online than a COBOL lie. It’s possible some SSA applications use 1875 as a placeholder (a choice made by programmers, not COBOL), but without official specs, that’s speculative. More likely, Musk’s team misread legacy data quirks—perhaps NULL fields, unverified birth dates, or survivor benefits tied to old records—as fraud. The real issue isn’t COBOL’s "lies" but aging systems struggling with data integrity, a problem modernization could address better than finger-pointing at a 60-year-old language.
Click to expand...
It JUST KEPT GETTING REPEATED, and mostly by people who evidently did not - or who stated they did not - know COBOL. I used COBOL for most of the nineties - we mostly got rid of it after Y2K - for a lot of reasons. I thought COBOL was a fantastic language for the kinds of processing we did, and if written well, was VERY READABLE. I loved the various kinds of data structuring - things like the 88 levels, where words and text could be substituted for specific data condtions. It was clearly best used for flat or plain indexed files instead of anything more complex - but most of the time, we never needed it.

But I digress - when you see people sneering and making references to computer languages they don't know - they're just repeating crap they saw elsewhere. You could question them about it and find out - they don't actually know it at all.

I've had my doubts about Elon's "team" as being young and brash and headstrong - such is youth, subsituting knowledge for wisdom - e.g. seeing fraud when ineptitude explains it better. But all I EVER SAW was conjecture about what MIGHT have happened, and assertion that that is what DID happen. And I find it very hard to believe that a team of extremely bright programmers who were examining the problem were just dumber than a handful of people NOT seeing it and guessing.

Then of course, I saw the distribution of ages - numbers from ages 120-129, 130-139 and so on into the 300's - and I realize, ok, there's really no logical way fraud is being conducted on a record saying someone is two hundred years old - but it does cast doubt on whether ANY of those records are being accurately maintained and audited. We often have test data that lists Mickey Mouse and Dorothy Gale and Clark Kent - but we KNOW it's bs data used for testing. This is REAL data, and it has crap in it.
 
You must log in or register to reply here.
Top